Lactation is associated with increased expression of bile acid transporters and an increased size and hydrophobicity of the bile acid pool in rats. ATP-binding cassette (ABC) transporters multidrug resistance protein 2 (Mdr2), Abcb11 [bile salt export pump (Bsep)], and Abcg5/Abcg8 heterodimers are essential for the biliary secretion of phospholipids, bile acids, and cholesterol, respectively. We investigated the expression of these transporters and secretion of their substrates in female control and lactating Sprague Dawley rats and C57BL/6 mice. Expression of Abcg5/Abcg8 mRNA was decreased by 97 and 60% by midlactation in rats and mice, respectively; protein levels of Abcg8 were below detection limits in lactating rats. Mdr2 mRNA expression was decreased in lactating rats and mice by 47 and 59%, respectively. Despite these changes in transporter expression, basal concentrations of cholesterol and phospholipid in bile were unchanged in rats and mice, whereas increased Bsep mRNA expression in early lactation coincided with an increased basal biliary bile acid concentration in lactating mice. Following taurocholate infusion, coupling of phospholipid and taurocholate secretion in bile of lactating mice was significantly impaired relative to control mice, with no significant changes in maximal secretion of cholesterol or bile acids. In rats, taurocholate infusion revealed a significantly impaired coupling of cholesterol to taurocholate secretion in bile in lactating vs. control animals. These data reveal marked utilization of an Abcg5/Abcg8-independent mechanism for basal biliary cholesterol secretion in rats during lactation, but a dependence on Abcg5/g8 for maximal biliary cholesterol secretion.

  • biliary secretory function
  • bile acids
  • phospholipid
  • taurocholate
  • multidrug resistance protein 2
  • bile salt export pump

the liver plays a central role in lipid and cholesterol homeostasis. Synthesis of bile acids from cholesterol and secretion of bile acids is highly integrated and tightly regulated to maintain cholesterol homeostasis, optimize intestinal lipid absorption, and minimize intracellular accumulation of cytotoxic bile acids (9). Bile acids, phospholipids, and cholesterol are secreted from hepatocytes and form mixed micelles that act in the small intestine to promote absorption of lipid-soluble nutrients, including dietary lipids, lipid-soluble vitamins, and cholesterol (19). During lactation in rats, bile flow, bile acid synthesis, and secretion are all increased (5, 21), as is the size and hydrophobicity of the bile acid pool (5, 43). We have shown that the expression of Na+/taurocholate cotransporting polypeptide (Ntcp) in the basolateral domain of the hepatocyte, the bile salt export pump (Bsep/Abcb11) expressed in the apical domain of the hepatocyte, and the ileal apical sodium-dependent bile acid transporter (Asbt) are all increased in the rat dam during lactation (7, 2527). We have further shown that the increase in the bile acid pool is mediated by increased expression of cholesterol 7α-hydroxylase (Cyp7a1), which catalyzes the rate-limiting step in the conversion of cholesterol to bile acids (43).

Lactation increases nutrient and energy demand in the rodent by four- and fivefold, resulting in a two- to threefold increase in food intake (11, 17). Although the efficiency of nutrient absorption in the dam increases to meet both the needs of the dam and for incorporation of nutrients into milk, lactating rats are likely in negative energy balance at all times of day (3, 41). We have speculated that the increased size and hydrophobicity of the bile acid pool might serve to increase the efficiency of absorption of cholesterol and lipid-soluble nutrients in lactation (7). An increase in the size and hydrophobicity of the bile acid pool would also likely impact the biliary excretion of lipids, i.e., phospholipids and cholesterol, into bile.

ATP-binding cassette (ABC) transporters multidrug resistance protein 2 (Mdr2), Bsep, and Abcg5/Abcg8 heterodimers in the canalicular membrane of hepatocytes mediate the biliary secretion of phospholipids, bile acids, and cholesterol, respectively. Mdr2 is a floppase that flops phosphatidylcholine from the inner to the outer leaflet of the canalicular membrane (34). Bsep is the primary bile acid transporter on the canalicular membrane and is responsible for biliary excretion of monovalent bile acids (2), whereas Abcg5 and Abcg8 form heterodimers that function to promote sterol excretion into bile (18).

Based on our data demonstrating increases in essential bile acid transporters in the liver and intestine during lactation (7, 2527), we hypothesized that the expression of the key transporters of mixed micelle components, i.e., Mdr2, Bsep, and Abcg5/Abcg8, would also be regulated during lactation. We measured mRNA expression of these transporters and basal concentrations of bile acids, phospholipids, and cholesterol in bile in rats and mice. We also infused taurocholate, a hydrophobic bile acid, to characterize the coupling of phospholipid and cholesterol biliary secretion to that of taurocholate. These studies demonstrated that, despite a nearly complete absence of expression of Abcg5/Abcg8 mRNA and Abcg8 protein in lactating rats, basal biliary concentrations of cholesterol were unchanged, consistent with an Abcg5/Abcg8-independent pathway of cholesterol secretion in lactating rats. However, maximal cholesterol secretion in bile was decreased in lactation following taurocholate infusion in the absence of Abcg5 and Abcg8. To determine if similar effects were present in other species, we characterized expression of Mdr2, Bsep, and Abcg5/Abcg8 and biliary excretion of their substrates in lactating mice. Whereas Mdr2 mRNA expression was decreased in both rats and mice, only mice exhibited a decrease in the maximal biliary secretion of phospholipids.


Chemicals and reagents.

General reagents were purchased from Sigma Aldrich (St. Louis, MO). TRIzol Reagent and Superscript III First-Strand Synthesis System for RT-PCR were purchased from Invitrogen Life Technologies (Carlsbad, CA). Light Cycler DNA Master SYBR Green 1, Light Cycler 480, and master mix/universal probes were purchased from Roche Diagnostics (Indianapolis, IN), and the RNeasy Mini kit was from Qiagen (Valencia, CA). Secondary antibodies and chemiluminescence reagents were purchased from Pierce Chemicals (Rockford, IL). Calnexin antibody was purchased from Nventa (San Diego, CA). Rabbit anti-Abcg5 and mouse anti-Abcg8 antibodies have been previously described (15, 45). Antibodies for immunofluorescence, including polyclonal antibody against ZO-1, Alexa Fluor-488-conjugated goat anti-mouse, Alexa Fluor-568-conjugated goat anti-rabbit, and ProLong gold antifade reagent with Dapi, were purchased from Invitrogen Life Technologies.

Animal care.

Female, 11- to 12-wk-old Sprague Dawley rats and C57BL/6 mice were purchased (Harlan Industries, Indianapolis, IN) as timed-pregnant or virgin controls and were maintained on a 12:12-h light-dark cycle (0600 lights on/1800 lights off) in a temperature-controlled environment. Animals had free access to Teklad Global Diet 2018 (Harlan Laboratories, Cincinnati, OH) and water. Litter size was culled within 24 h of birth to 9–11 or 7–9 pups for rats and mice, respectively. Animal protocols were conducted in accordance with the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee of the University of Kentucky. Female virgin rats or mice (control group), rats at late gestation (day 19 of gestation), and postpartum (PP) rats or mice during early lactation (day 3 and 5 PP), midlactation (day 10 and 12 PP), and late lactation (day 19 and 20 PP, just before weaning) were killed at 1600. Postweaning rats were killed 55 days following weaning of the pups.

Real-time PCR analysis.

Total RNA from liver was isolated with TRIzol Reagent and purified with the Qiagen RNeasy Mini kit, followed by cDNA synthesis with SuperScript III. To determine mRNA expression, real-time PCR was performed on a Roche LightCycler with a SYBR Green kit (rat Abcg5, Abcg8, Mdr2, 18S) and the Light Cycler480 with Light Cycler 480 master mix/universal probes (mouse Abcg5, Abcg8, Mdr2, Bsep, 18S). Amplification of diluted cDNA template was used to create a semiquantitative standard curve by plotting the cycle number vs. the log of the fluorescence measurement at the threshold. Conditions used for amplification using the LightCycler were as follows: denaturation for 30 s at 95°C and 40 cycles of 95°C for 0 s; 60°C for 15 s (Abcg5), 65°C for 15 s (Abcg8), 57°C for 15 s (Mdr2), 56°C for 20 s (18S); and 72°C for 15 s (Abcg5, Abcg8, Mdr2) and 30 s (18S). Primers were as follows: Abcg5 (forward, 5′-CGCAGGAACCGCATTGAAA-3′; reverse, 5′-TGTCGAAGTGGTGGAAGAGCT-3′); Abcg8 (forward, 5′-GATGCTGGCTATCATAGGGAGC-3′; reverse, 5′-TCTCTGCCTGTGATAACGTCGA-3′); Mdr2 (forward, 5′-CCCACAGAGGGTACGATTAGCA-3′; reverse, 5′-CGCCGATGAATTCCCTTAGAC-3′); and 18S (forward, 5′-GTAACCCGTTGAACCCCATT-3′; reverse, 5′-CCATCCAATCGGTAGTAGCG-3′). For the LightCycler 480, primers and probe sets were designed using the Roche Universal Probe Library (www.universalprobelibrary.com) to amplify intronic-spanning regions for the gene. Reactions were performed as follows: denaturation for 5 min at 95°C, and 45 cycles of 95°C for 15 s; 60°C for 30 s. The following probes were used: no. 81 (18S), no. 10 (Bsep and Abcg8), no. 31 (Abcg5), and no. 4 (Mdr2).

Preparation of membrane proteins.

Whole liver (mouse) or 100–200 mg liver (rat) were homogenized using a Polytron homogenizer in buffer A (250 mM sucrose, 2 mM MgCl2, and 20 mM Tris·HCl, pH 7.5) and centrifuged at 2,000 g for 10 min at 4°C. The supernatant was collected and centrifuged at 120,000 g for 45 min at 4°C. The membrane pellet was suspended in buffer B (80 mM NaCl, 2 mM CaCl2, 1% Triton X-100, and 50 mM Tris·HCl, pH 8), and protein concentrations were determined by BCA assay (Pierce).

Immunoblot analysis.

Protein samples were prepared in protein sample buffer containing β-mercaptoethanol (final concentration 1.2%) and were boiled at 95°C for 5 min. Size fractionation was performed on 10% SDS-polyacrylamide gel and transferred to nitrocellulose membranes. Membranes were incubated in blocking buffer [5% (wt/vol) dry milk in Tris-buffered saline with 0.2% Tween 20] for 1 h at 4°C. Primary antibodies were diluted 1:500 in blocking buffer and incubated overnight at 4°C. Horseradish peroxidase-conjugated goat anti-rabbit IgG and goat anti-mouse IgG were diluted (1:15,000) and incubated for 1 h at room temperature. Proteins were visualized by chemiluminescence, and protein loading was assessed by blotting with calnexin. Densitometry was used for semiquantitative analysis of expression levels.

Confocal microscopy analysis of Abcg8 and ZO-1.

Livers from virgin and lactating Sprague Dawley rats [PP day (d) 21] were dissected, immediately embedded in Tissue-Tek Optimum Cutting Temperature Compound 4583 (Sakura Finetek), placed on dry ice, and then stored at −80°C. Liver sections (14 μm) were prepared with a Microm HM 560 CryoStar microtome. Sections were fixed with methanol (for 5 min at −20°C), rehydrated in PBS (pH 7.4), and incubated in buffer A [PBS + 1% BSA (wt/vol)] at 22°C for 1 h to block nonspecific binding. For Abcg8 and ZO-1, sections were incubated with a monoclonal antibody against Abcg8 (16 μg/ml) and a polyclonal antibody against ZO-1 at a dilution of 1:100 in buffer A at 4°C for 16 h. Sections were washed with buffer A and incubated with Alexa Fluor-488-conjugated goat anti-mouse and Alexa Fluor-568-conjugated goat anti-rabbit (2 μg/ml) in buffer A at 22°C for 1 h. After being washed three times with PBS and one time with distilled water, slices were mounted in aqueous mounting media (ProLong gold antifade reagent with Dapi) and imaged on a Zeiss Axiovert 200M confocal microscope equipped with an Apotome.

Basal bile collection.

Rats and mice were anesthetized with urethane (1 g/kg ip), and the bile duct was cannulated (rats) or ligated and the gall bladder cannulated (mice) with PE-10 tubing. In rats, basal bile was collected for ∼10 min before single-pass liver perfusion following cannulation of the bile duct. In mice, bile was collected for 45 min. For both rats and mice, bile volume was determined gravimetrically assuming a density of 1.0. Samples were frozen at −20°C until analysis of bile components.

Single-pass liver perfusion.

Single-pass rat liver perfusion was performed as previously described (26). The liver was perfused at a flow rate of 3.5–4.0 ml·min−1·g−1 via the portal vein with Krebs-Henseleit buffer (in mM: 118.5 NaCl, 24.9 NaHCO3, 1.2 KH2PO4, 1.19 MgSO4, 4.74 KCl, 1.27 CaCl2, and 5 glucose, pH 7.4). Perfusate was oxygenated with 95% O2-5% CO2, and the liver was maintained at 36 ± 1°C. Bile was collected every 5 min, and volume was determined gravimetrically assuming a density of 1.0. Taurocholate was infused in the portal vein cannula at increasing concentrations ranging from 4 to 250 nmol/ml for 15 min each. In mice, the liver was perfused through the portal vein with Krebs-Henseleit buffer at a flow rate of 5 ml·min−1·g−1 liver and bile collected every 10 min. Taurocholate was infused in the portal vein cannula at increasing concentrations ranging from 5 to 120 nmol/ml for 10 min each.

Total phospholipids and cholesterol were measured in bile using enzymatic colorimetric kits from Wako Pure Chemical Industries (Richmond, VA). Quantitation of total bile acids in bile was performed enzymatically by measuring 3α-hydroxy bile acids as described previously (39) for rats or using a colorimetric kit from Wako for mice.

Statistical analysis.

All data are expressed as means ± SE for n = 3–8 animals/group. Statistical analysis was performed with Student's t-test, one way-ANOVA followed by Tukey's multiple comparison test, or linear regression analysis with GraphPad 4.0 software (San Diego, CA) as indicated in the legends for Figs. 16.

Fig. 1.

Expression of ATP-binding cassette (Abc) transporters Abcg5, Abcg8, and multidrug resistance protein 2 (Mdr2) in control (C), late pregnancy (G19), postpartum [PP days (d) 3, 10, 21], and 55 days postlactating (PL) d55 rats. Real-time PCR was performed in duplicate on cDNA synthesized from liver total RNA (n = 3–6) to amplify Abcg5 (A), Abcg8 (B), and Mdr2 (D) mRNA expression. Each bar represents the mean ± SE. C: immunoblot analysis was performed to determine protein expression of Abcg8 in control and PPd10 rats (ns, nonspecific). *P < 0.05, **P < 0.01, and *** P < 0.003 vs. control as indicated by Student's unpaired t-test or Mann Whitney test (Abcg5, C vs. PPd10).

Fig. 2.

Immunolocalization of Abcg8 in rat liver sections from lactating and virgin rats. Cryosections (14 μm) from liver of lactating (A-C) and virgin (D-F) rats were probed with antibodies directed against Abcg8 (A and D) and the tight junction marker ZO-1 (B and E). Nuclei were labeled with DAPI (blue). Merged images are shown in C and F. Scale bar is equal to 20 μm.

Fig. 3.

Effect of taurocholate (TC) infusion (4–200 nmol/ml) on biliary cholesterol (CH) and phospholipid (PL) secretion in control and PPd11–12 rats. Each regression line represents bile acid (BA)-dependent PL secretion (A), BA-dependent CH secretion (B), and PL-dependent CH secretion (C) in control (solid line, n = 7) and PPd12 (broken line, n = 10) rats. Data points represent individual samples of control (■) and PPd12 (▵) rats. Regression analysis results are shown in Table 2.

Fig. 4.

Expression of Bsep and Mdr2 in control and PPd5, -12, and -20 mice. Real-time PCR was performed in duplicate on cDNA synthesized from liver total RNA of control and postpartum mice, n = 4–8/group for Bsep in mice (A) or Mdr2 (B). Each bar represents the mean ± SE. *P < 0.05 and **P < 0.01 vs. control, as indicated by one-way ANOVA.

Fig. 5.

Expression of Abcg5 and Abcg8 in control and lactating (PPd5, -12, and -20) mice. Real-time PCR was performed in duplicate on cDNA synthesized from liver total RNA in control and postpartum animals (n = 4–7) to amplify Abcg5 (A) and Abcg8 (B) mRNA expression. Each bar represents the mean ± SE. Immunoblot analysis was performed to determine protein expression of Abcg5 (C) and Abcg8 (D). *P < 0.05, **P < 0.01, and ***P < 0.001 vs. control mice as indicated by one-way ANOVA. Immunoblot analysis of Abcg5 and Abcg8 showed no significant difference in protein expression between C and PP mice despite changes in mRNA expression.

Fig. 6.

Influence of TC infusion (5–120 nmol/ml) on CH and PL secretion in control and lactating (PPd12) mice. Each regression line represents BA-dependent PL secretion (A), BA-dependent CH secretion (B), and PL-dependent CH secretion (C) in control (solid line, n = 6) and PPd12 (broken line, n = 4) mice. Data points represent individual samples of control (■) and PPd12 (▵) mice. Results of the regression analysis are shown in Table 2.


Changes in expression of transporters of mixed micelle components in rats.

Expression of bile acid transporters in the basolateral domain (Ntcp) and the canalicular domain (Bsep) of the hepatocyte, and the apical domain of the enterocyte (Asbt) are all increased in early (Ntcp and Bsep) and midlate (Asbt) lactation in the rat (7, 2527). Furthermore, the bile acid pool size is increased by midlactation (PPd10) in rats and remains elevated until weaning (7, 43). We therefore extended these studies to determine if there were changes in the expression of Mdr2 and Abcg5/Abcg8 that mediate biliary secretion of phospholipids and cholesterol, respectively. Mdr2 mRNA expression was significantly decreased by 45% at PPd10 (Fig. 1D), whereas expression of Abcg5 and Abcg8 mRNA was profoundly decreased throughout lactation (Fig. 1, A–C). Thus Abcg5 mRNA expression was decreased 80% at PPd3, 97% at PPd10, and 93% at PPd21. Similarly, Abcg8 mRNA expression was decreased 94% at PPd3, 98% at PPd10, and 97% at PPd21 (Fig. 1, A and B). Pregnant or postweaning rats (day 55) showed no significant difference in Abcg5 or Abcg8 mRNA expression compared with controls (Fig. 1, A and B). Immunoblot analysis confirmed the absence of Abcg8 expression in PPd10 rats (Fig. 1C). Confocal immunofluorescence analysis showed that while ZO-1 localization in the apical membrane and architecture were maintained, Abcg8 expression could not be detected in hepatocytes from lactating rats at PPd21 (Fig. 2). However, there were no differences in bile acid, phospholipid, and cholesterol concentrations in bile under basal conditions (Table 1).

View this table:
Table 1.

Basal bile concentrations of bile acids, phospholipids, and cholesterol in control vs. lactating rats and mice

Effects of taurocholate infusion on secretion of mixed micelle components in rats.

The biliary excretion of phospholipids and cholesterol are known to be coupled to that of bile acids (18) such that taurocholate infusion increases secretion of bile acids (7), which in turn increases incorporation of phospholipids and cholesterol into bile acid micelles, thereby increasing their secretion in bile (33). Because there were no major changes in biliary bile acid, phospholipid, or cholesterol concentrations under basal conditions, we measured the coupling of cholesterol and phospholipids to the increased bile acid secretion following infusion of taurocholate. Secretion rates were normalized to liver weight, and regression analysis was used to model secretion of phospholipids and cholesterol as a function of bile acid secretion (Fig. 3, A and B) or cholesterol secretion as a function of phospholipid secretion (Fig. 3C). The slopes, which denote the molecular coupling of the two substrates, were all significantly different from zero. As shown in Fig. 3 and Table 2, the coupling of phospholipids to bile acid secretion was only modestly impaired (P < 0.005), despite the decrease in Mdr2 mRNA expression seen at this time. In contrast, cholesterol secretion was markedly decreased in response to taurocholate infusion, as shown by the decreased slope of the regression line, consistent with a decreased coupling of cholesterol to bile acid secretion in bile (P < 0.0001; Fig. 3B and Table 2). Although the slopes were unchanged between control and postpartum animals for phospholipid-dependent cholesterol secretion, postpartum animals had visibly lower secretion rates of cholesterol (Fig. 3C) with a significantly lower positive intercept (0.3741, control; 0.03284, PPd10; P < 0.0001).

View this table:
Table 2.

Influence of taurocholate on bile secretion in control vs. lactating rats and mice

Changes in expression of transporters of mixed micelle components and their substrates in mice.

To determine if the markedly decreased expression of Abcg5/Abcg8 seen in the lactating rat was also present in lactating mice, we examined expression of Bsep, Mdr2, and Abcg5/Abcg8 in lactating mice and female controls. Bsep mRNA expression was significantly elevated during early lactation (2.5-fold increase, PPd5; Fig. 4A), but there were no differences compared with controls at later stages of lactation. The mouse also showed a significant 58% decrease in Mdr2 mRNA expression at midlactation (PPd12) (Fig. 4B) but only a moderately decreased mRNA expression of Abcg5 and Abcg8. Thus Abcg5 mRNA expression was decreased 40% at PPd5, 65% at PPd12, and 53% at PPd20 (Fig. 5A), whereas Abcg8 mRNA expression was not significantly different at PPd5 but was decreased 63% at PPd12 and 53% at PPd20 (Fig. 5B). Although protein expression of Abcg5 and Abcg8 was somewhat decreased (Abcg5 by 28% and Abcg8 by 21%) at PPd12, these decreases were not significantly different compared with controls (Fig. 5C).

Next, we measured the basal concentrations of cholesterol, phospholipid, and bile acids in bile to determine if altered transporter expression correlated with transport activity. In mice, the bile acid concentration was significantly elevated 1.5-fold in PPd11 mice (Table 1), consistent with increased expression of Bsep mRNA. In agreement with no significant changes in mouse Abcg5 and Abcg8 protein levels, there were no changes in biliary cholesterol concentrations in the postpartum mouse. Phospholipid concentrations were also not changed in postpartum mice compared with controls (Table 1), despite decreases in Mdr2 mRNA expression (Fig. 4B).

Effects of taurocholate infusion on secretion of mixed micelle components in mice.

There was significant positive coupling between bile acids and phospholipid secretion in both control and postpartum mice; however, lactating mice showed a decreased slope (P < 0.005; Fig. 6A and Table 2), consistent with the decreased Mdr2 mRNA expression in mice at this time (Fig. 4B). Bile acid-dependent cholesterol secretion remained positively coupled in control and lactating mice to essentially the same extent (Fig. 6B). Phospholipid-dependent cholesterol secretion was also positively coupled (Fig. 6C), and postpartum animals had an increased slope (P < 0.0001; Table 2), consistent with the decreased coupling of taurocholate and phospholipid biliary secretion.


The most striking and unexpected findings in the present studies were the profoundly decreased expression of Abcg5/Abcg8 in lactating rats, particularly in the face of unchanged basal cholesterol concentrations in bile, compared with control rats. Earlier studies (20) had shown an increased cholesterol concentration in bile of lactating rats. Significant evidence indicates that the half-transporters Abcg5 and Abcg8 are required for cholesterol secretion in bile, in that deletion of both half-transporters in mice leads to a marked decrease (90%) in biliary cholesterol (45). Furthermore, cholesterol secretion is linearly correlated with the gene copy number of Abcg5/Abcg8 in mice (44). In humans, mutations in either ABCG5 or ABCG8 gene cause an autosomal recessive disorder known as sitosterolemia, characterized by decreased biliary excretion of cholesterol, and increased intestinal absorption of cholesterol and phytosterols, such as sitosterol (4). However, there is also evidence of Abcg5/Abcg8-independent pathways of cholesterol secretion in mice. Diosgenin is a plant sterol that stimulates cholesterol excretion (30). This induction of cholesterol secretion is independent of Abcg5/Abcg8 induction (22), even though biliary cholesterol secretion is not induced in Abcg8−/− mice fed diosgenin, suggesting that expression of the cotransporters is still required.

The present study offers further evidence in a physiological model of an Abcg5/Abcg8-independent pathway of cholesterol secretion in lactating rats. However, consistent with the requirement for Abcg5/Abcg8, there was a significantly decreased coupling of cholesterol secretion to bile acid secretion in bile in lactating vs. control rats upon infusion of the relatively hydrophobic bile salt taurocholate. Although the basis for the normal basal cholesterol secretion in lactating rats is not clear, it is likely because of the increased size and hydrophobicity of the bile acid pool. These changes, coupled with the increased flux of these hydrophobic bile acids across the canalicular membrane in lactating rats, is sufficient to extract adequate amounts of cholesterol from the membrane to maintain basal concentrations without the need for a transporter.

Transport of cholesterol has been demonstrated in ABC transporters besides Abcg5/Abcg8 in the presence of an extracellular acceptor (28, 29). Thus, ABCA1 transports cholesterol when either apolipoprotein A-I or taurocholate are present to serve as acceptor molecules (29). Similarly, Mdr2 expressed in HEK cells effluxed cholesterol together with phospholipid when 0.5–1 mM taurocholate was present in the media as an acceptor molecule (28). These data clearly demonstrate that Mdr2 is able to directly efflux cholesterol together with phospholipid from cells as long as there is an acceptor molecule, in this case, taurocholate (28). Our data indicate, however, that, in the presence of a marked increase in the flux of infused taurocholate across the canalicular membrane, there is apparently insufficient cholesterol in the outer hemi-leaflet of the membrane bilayer to mediate sustained high levels of biliary cholesterol in the absence of transporters to flop additional cholesterol from the inner to the outer leaflet of the membrane. Evidence for such a mechanism is also seen in Atp8b1-deficient mice. Mutations in mice deficient in both Atp8b1 and Abcg8 demonstrate increased cholesterol secretion compared with wild-type controls; this has been attributed to decreased detergent resistance of the canalicular membrane because of the accumulation of phosphatidylserine in the outer leaflet of the membrane (16). However, we found no changes in mRNA expression of Atp8b1 in control vs. lactating rats (data not shown). Decreased detergent resistance may also play a part in lactation when bile acids, which act as detergents, are increased. Finally, there may be an alternative transporter for cholesterol when Abcg5 and Abcg8 are not present. Further studies would be needed to identify the mediators of such a putative pathway of cholesterol secretion. Although scavenger receptor class B type I has been suggested to mediate cholesterol flux across the membrane (42), we found no changes in its mRNA expression in lactating vs. control rats (data not shown).

The physiological or teleological bases for loss of expression of Abcg5/Abcg8 in lactation are not known. We postulate that the increased size and hydrophobicity of the bile acid pool in lactating rats relative to control rats would markedly increase the basal levels of cholesterol in bile were normal expression of Abcg5/Abcg8 retained. Thus the relatively hydrophilic muricholic acids that are the major bile acids in control rats (75 mol/100 mol of the bile acid pool) are decreased to 62 mol/100 mol by 14 days of lactation, whereas the relatively hydrophobic taurocholic acid is increased from 20 mol/100 mol in control rats to 31 mol/100 mol in lactating rats (43). Loss of biliary cholesterol could be detrimental for several reasons. Most importantly, cholesterol and lipids are essential components in milk that support membrane synthesis and brain development in the pups. Cholesterol is secreted in milk [∼16 mg/day in the rat (10)], with 32–40% synthesized de novo in the mammary gland and 11% from dietary sources so that cholesterol synthesized in the liver makes up the majority of cholesterol secreted in milk (14). Also, biliary excretion of cholesterol serves as a route of its elimination from the body, both directly and following its catabolism to bile acids. The liver obtains cholesterol from peripheral tissues, intestinal absorption, and de novo synthesis. Fifty percent of cholesterol catabolized in the liver is used for the production of bile acids, and 40% is secreted directly in bile (8). Hepatic cholesterol synthesis increases ∼50 and 300%/total organ at 14 and 21 days of lactation, respectively (13). Because the bile acid pool size is increased almost threefold in lactation, this implies that >50% of cholesterol is catabolized to bile acids in lactation, mediated by the increased expression of Cyp7al, the rate-limiting enzyme in this process. Finally, cholesterol is a component of mixed micelles in bile (19). These mixed micelles are essential for emulsifying lipid-soluble nutrients for absorption from the small intestine. Increased cholesterol secretion in bile decreases cholesterol absorption in the small intestine (23). Taken together, these data suggest that the lactating dam needs to conserve cholesterol; the decreased expression of Abcg5 and Abcg8 in the liver during lactation may thus serve to minimize cholesterol elimination in bile.

Hormonal changes that occur during lactation may play a role in Abcg5 and Abcg8 regulation. We have shown that prolactin, one of the essential hormones involved in maintaining milk production, is involved in regulation of biliary secretory function. Prolactin increases the capacity of the liver to secrete taurocholate in bile (26) by increasing the expression of Ntcp and Bsep in the rat during lactation (7). Leptin is another hormone that regulates biliary secretory function. Leptin is secreted from adipocytes and increases energy expenditure while reducing energy intake by decreasing appetite (1). Lactation is characterized by hypoleptinemia, likely reflecting the required increased energy intake (35, 38). Leptin also promotes hepatic cholesterol clearance (40), and mice lacking leptin or the leptin receptor (ob/ob and db/db, respectively) have reduced protein expression of Abcg5 and Abcg8 (36). In addition, leptin administration to ob/ob mice decreases the size and hydrophobicity of the bile acid pool (20). Decreased leptin during lactation may thereby contribute to decrease the maximal cholesterol excretion and permit expansion of the bile acid pool. Prolactin has been reported to inhibit leptin release from adipocytes (6, 24) and may thereby play a role in leptin regulation during lactation.

Mdr2 is required for Abcg5- and Abcg8-dependent biliary cholesterol secretion, where transgenic mice that overexpress ABCG5/ABCG8, but lack Mdr2, also lack increased biliary cholesterol secretion (23). Phospholipid secretion in bile is abolished in Mdr2−/− mice, and cholesterol secretion is significantly inhibited; however, Mdr2+/− animals had biliary cholesterol concentrations similar to controls (31, 32, 37). Although we found Mdr2 mRNA expression to be decreased in both rats and mice at midlactation, mRNA may not reflect protein levels, since Mdr2 protein has been reported to increase in the rat during lactation (38).

Clear species differences were seen in Abcg5/Abcg8 expression and cholesterol secretion in bile upon infusion of taurocholate. Although Abcg5/Abcg8 expression in the rat was essentially completely abolished during lactation at the mRNA and protein level (Fig. 1), mice showed a modest decrease in mRNA expression, and protein levels were not affected (Fig. 5). Species and gender differences in Abcg5/Abcg8 expression and regulation have been shown between rats and mice, where male rats tend to have higher expression in the liver than females, and, upon cholesterol feeding, Abcg5/Abcg8 mRNA expression decreases in rats but increases in mice (12). In the present studies, lactating rats were able to maintain maximal phospholipid secretion (Fig. 3), whereas lactating mice had significantly reduced phospholipid secretion in response to infusion of increasing concentrations of taurocholate (Fig. 6A). Although mRNA expression may not reflect protein levels, there was a positive correlation between Mdr2 mRNA expression and maximal phospholipid secretion in mice.

In summary, the present studies demonstrate an Abcg5- and Abcg8-independent pathway of sustained basal cholesterol biliary concentration and secretion in the lactating rat. These findings represent a novel yet physiological model in which the lactating dam is able to downregulate expression of these important transporters in the face of an increased bile acid pool size that is more hydrophobic, apparently to minimize loss of biliary cholesterol.


This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grants DK-46293 (to M. Vore) and 1R01DK-080874 (to G. Graf) and National Institute of Environmental Health Sciences Training Grant T-32ES-07266 (D. Coy and C. R. Wooton-Kee).


No conflicts of interest are declared by the authors.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
View Abstract